Chapter 9

Marketing Analytics
Based on First Principles :
Chapter 9
Using Customer Lifetime Value

for Customer Selection
© Palmatier, Petersen, and Germann 1

Agenda
 Learning Objectives
 Introduction
 Unprofitable Customers
 The Idea behind Customer Lifetime Value (CLV)
 CLV in Action
 Customer Lifetime Value (CLV)

 An Example
 Using CLV to Determine Acquisition Costs of Customers
 Using CLV for Prospecting Decisions
 Estimating the Retention Rate
 Linking CLV to Firm Value
 Some Additional Thoughts
 Summary
 Takeaways

Learning Objectives
 Know the idea behind Customer Lifetime Value (CLV).
 Understand the “standard” CLV formula as well as the “alternative” CLV

formula. Be able to explain the difference between the two formulas.
 Describe how to use CLV for prospecting decisions.
 Know how to use CLV to determine maximum acquisition costs.
 Explain how to do a return on marketing investment (ROMI) calculation

based on the findings from a CLV analysis.
 Understand the different approaches researchers take to estimate the

retention rate.
 Know how CLV is linked to firm value.

Agenda
 Introduction
 CLV in Action

 An Example
 Summary
 Takeaways

Unprofitable Customers
Not all customers are equally profitable.

 The Pareto Principle, also known as the 80/20 rule, is often relevant when
looking at the distribution of your customers’ Customer Lifetime Values
(CLVs).
 Usually about 10-20% of your customers help to generate about 80-90% of

your total profit.
 Further, the next 40-60% of your customers tend to be a little above

breakeven.
 That means that the remainder of your customers are usually unprofitable.

 If a company cannot turn unprofitable customers into profitable ones, the

company may want to dismiss or find less expensive ways to manage those
customers who cost more to serve than they are worth.
 Some suggest companies should fire unprofitable customers. Doing so can

give companies the bandwidth and resources to take care of the profitable
customers. But, be careful…

 Consider the following example:

 In June 2007, Sprint Nextel sent out letters to about 1,000 customers informing
them that their contracts had been terminated. For about a year, Sprint had been
tracking the calls made by these customers. According to a Sprint Nextel
spokesperson, some of them called customer care hundreds of times a month.
Eventually, Sprint Nextel determined it could not meet the needs of these
customers and, therefore, cut off their service (Sprint Nextel also waived any
cancelation fees).
 Sprint Nextel received quite a bit of negative word of mouth (WOM) when it fired
the 1,000 customers, and its stock price dropped by about 2% around the time
when the announcement was made (Srivastava 2007).
 However, it can be possible to end relationships with customers and come

out ahead. For instance, when Allstate Insurance dropped 95,000
homeowner policies in Florida, its stock price surged 5.7%.

The Idea behind Customer Lifetime Value
(CLV)
 One of the most widely used metrics for determining how much a customer
is worth to a company is customer lifetime value, or CLV.
 The idea behind CLV is fairly intuitive.

 A company engages in some efforts to win a customer.
 Once a prospect becomes a customer, the company expects to earn some revenues
from that customer.
 Serving the customer will also likely come at a cost to the company. Indeed, it will
cost the company something to make and supply the product (be it a good or a
service) and to retain the customer.
 The customer’s expected retention rate, or how likely it is that the customer will
stay with the company, is taken into account in a CLV analysis.
 Moreover, the customer’s future net cash flows need to be discounted to today’s
value, which a CLV analysis again does.

CLV in Action
 As customers, most of us have individual CLV scores with at least some

companies we interact with.
 “Your [CLV] score can determine the prices you pay, the products and ads
you see and the perks you receive. Credit card companies use the scoring
systems to decide what to offer customers who want to cancel their cards.
Wireless carriers route high-value callers immediately to their most
skilled agents. At some airlines, a high score increases the odds of a seat
upgrade.”
– Safdar (from the Wall Street Journal in 2018)
 “Not all customers deserve a company’s best efforts”, so whether you are
on hold for 50 minutes or 1 minute when calling a customer service
hotline might be because of your (secret) CLV score.
– Marketing Professor Pete Fader

Agenda
 Introduction
 CLV in Action

 An Example
 Summary
 Takeaways

An Example
 Customer Lifetime Value (CLV) is the net present value of all future
streams of profits that an individual customer generates over the life of
his/her business with the company.
 For example, Nick likes to buy a new snowboard at the beginning of every
snowboarding season. This year, he bought an Asher snowboard for which
he paid $500. Let us assume that it cost Asher $150 (i.e., variable costs) to
make the snowboard, so Asher’s gross margin on Nick’s purchase was
$350, i.e.:
Gross Margin ($M) = Price – Unit Variable Cost
Asher $M = $500 – $150 = $350

An Example
 As Nick likes to purchase a new snowboard every year, Asher could calculate his
CLV as follows:
+ = $350 + $350 + $350 .....
Year 1 Year 2 Year 3
 However, as any introduction to finance class teaches, the $350 Asher will earn
from Nick’s future purchases are not worth the same as the $350 it earns today.
Instead, it is important to account for the time value of money. Thus, we need to
discount the earnings as follows:
+ + + .....
where d is the discount rate Asher choses (let us assume Asher uses a discount rate of
10%).

An Example
 Another aspect that Asher needs to consider is that Nick’s loyalty toward
Asher may diminish over time. In other words, Asher needs to recognize
that Nick’s repeat business is not guaranteed and hence account for the
probability that Nick will still be a customer in time period t. This can be
done as follows:
+ + + .....
where p is the probability that Nick will buy in a given time period (which in
turn is derived from r = retention rate).

An Example
 Let us assume that there is a 90% probability that Nick will make a purchase in time
period t, given that he bought in time period t-1. Thus, the retention rate r = .90 or 90%,
and the unconditional purchase probabilities per time period are:
Time period 1 = = .90
Time period 2 = x = .90 x .90 = 0.81
Time period 3 = x x = .90 x .90 x .90 = 0.729
 If the retention rate is constant (as is the case here), we can represent the unconditional
purchase probability as = rt. Thus, assuming a constant retention rate, Asher can
calculate Nick’s CLV as follows:
+ + + .....
or

An Example
 A question that frequently arises is if the CLV would not get very large
since we are adding up time periods from 1 to infinity (i.e., stands for
infinity).
 However, due to the discounting of the future earnings as well as the
increasing chance over time that the customer will churn (i.e., exit the
relationship), this is often not an issue.
 In fact, if we assume that the gross margin ($M) is the same in each time
period, Nick’s CLV can be converted from an infinite sum into the
following formula:
= $M

An Example
 The formula is the “standard” CLV formula which assumes that a customer’s margin and
retention rate do not vary over time. Thus, using this formula, Nick’s CLV is:
$350 = $1,575
 An alternative CLV formula that is also frequently used assumes that the initial cash flow
(i.e., $M) is certain and received at the beginning of the first period. Using this alternative
formula, Nick’s CLV is calculated as follows:
+ + + .....
or
= $M
 Using this alternative CLV formula, Nick’s CLV is:
$350 = $1,925

Using CLV to Determine Acquisition Costs of
Customers
 How much should they spend on acquiring customers like Nick? The CLV
formulas can be used to answer this question. Let $AA be the average
acquisition cost for a new customer. The goal would be that:
or
- $AA
or
- $AA
 We use the alternative CLV formula here (indicated by CLVa). Thus, Asher
should spend, at most, $1,925 to acquire customers such as Nick.

Using CLV for Prospecting Decisions
 Not all prospects will become customers. Instead, only a percentage of prospects (if at
all) will become customers. Thus, the formula has to be further modified to understand
how much a company should spend on its customer acquisition efforts, i.e., its
acquisition efforts on prospects overall and not just individual customers such as Nick.
This can be done by calculating a prospect’s lifetime value (or PLV).
 Let us say “$A” is the amount of acquisition spending per prospect. Moreover, let “a” be
the probability of a prospect becoming a customer, i.e., a = the acquisition rate. Thus, the
goal would be that:
or
- $A
or
- $A

 Hence, a prospect’s lifetime value (PLV) can be calculated as follows:
PLV = - $A
 Let us assume Asher’s Marketers are considering spending $500,000 on a new

branded snowboarding video they plan to distribute via YouTube. They estimate
the video will reach 1,000,000 snowboarders.
 Asher’s Marketers expect the video will convince 0.1% of the viewers to purchase
a new Asher snowboard (with a gross margin $M of $350) every year for the
foreseeable future (like Nick).
 Let us also assume the CLV of the acquired prospects is $1,925 (as calculated above
for Nick). Also, $A = $500,000 / 1,000,000 = $0.5. To see if the YouTube video is
economically attractive, Asher’s Marketers use the formula.
 Thus, the PLV of each of the 1,000,000 viewers is:
PLV = 0.001 x $1,925 - $0.5 = $1.925 - $0.5 = $1.425 ≈ $1.43

 Thus, the expected lifetime value of a prospect is approximately $1.43. This is the most Asher
would want to spend on the video campaign per prospect. Moreover, the total expected value of
the prospecting effort (i.e., the branded snowboarding video) is 1,000,000 * $1.43 = $1,430,000.
Hence, at the most, this is how much Asher would want to spend on the video campaign in total.
 Further, the total expected value can also be used to calculate the return on marketing
investment (ROMI) of the YouTube video based on the following formula for ROMI.
 Using the ROMI formula, the video campaign’s ROMI is (($1,430,000 - $500,000) / $500,000) =
1.86 or 186%.
 Asher’s Marketers might be uncertain about the acquisition rate “a”, i.e., the probability of a
prospect becoming a customer. Therefore, another way to approach the problem is to ask what
the acquisition rate must be to break even with the campaign. Given the above equation and
numbers, they can calculate the break-even acquisition rate as follows:
=

 Thus, as long as Asher’s Marketers believe more than 0.026% of the

prospects will be acquired as customers and the average customer lifetime
value of a new Asher customer is $1,925, the video will have a positive
economic impact.
 That said, measuring the incremental value provided by a campaign can be

difficult. Using CLV may be one way to measure this value. If one is able to
measure the CLV of each targeted customer before and after the marketing
campaign, one could use the following equation to measure the
incremental value produced by the campaign:
 If marketers do not have a CLV for each customer before the start of a
marketing campaign, they may have to make some assumptions about
what each customer would have done (i.e., would that customer have
purchased anyway?) had the campaign not been run.
Estimating the Retention Rate
 Calculating CLV requires an estimate of each customer’s retention rate. So

how should one go about estimating retention rate?
 In short, firms use many different approaches to estimate retention rate.

 Although not ideal, some managers simply guess at it.
 Other managers estimate retention rate by considering the customer’s
transaction history.
 Further, some managers use a customer’s past transaction history as well as
other potential drivers of repurchase behavior (e.g., the firm’s marketing efforts
or the customer’s interactions with the firm’s website) – think Logistic
Regression from Chapter 8.
 As another example, some managers may calculate the probability of a

customer being alive in a certain period (P(alive)) given that customer’s
past purchasing behavior.

 Calculating P(Alive)
 Let’s assume a customer made four purchases from a firm this year
(January – December) and that that customer’s last purchase from the
firm was during the month of October. We are now at the end of December
and we want to estimate the customer’s probability of being alive (i.e., still
active in the relationship with the firm) in January. We can estimate that
probability using the following formula:
where P(alive) is the probability of the customer making a purchase from

the firm during the period of interest (i.e., in January), t is the time of the
last purchase and n is the number of purchases in a given period (here
January – December).
 Calculating P(Alive)
t is calculated as 10/12 = 0.833
(10 because the last purchase was in October [the 10th month of the period]
and 12 because we are interested in the probability of a purchase at the
start of January [i.e., right after the 12th month of the period])
n=4
because the customer made four purchases during the focal period (January
– December).
Thus, P(alive) = 0.833^4 = 0.48 or 48%. This probability, in turn, can then be
used as a customer-specific retention rate.

Linking CLV to Firm Value
 Once a firm has estimated the CLV score of all its current and prospective
customers, it can estimate its total customer lifetime value (or total
customer equity).
 To increase the total customer lifetime value, firms have three levers they
can use:
 Acquiring more profitable customers (or re-balancing the customer mix to focus
on the customers with the greatest return) and firing the unprofitable ones.
 Increasing customer retention rate among the profitable customers.
 Developing existing customers by trading them up and across the product
portfolio to increase their sales and profits.

Linking CLV to Firm Value
 However, there are three important reasons why customer retention –

especially the retention of customers with a positive CLV – is especially
critical to the success of a firm (i.e., firm value).
 Acquiring customers can cost significantly more than retaining customers.
 The customer profit rate tends to increase over the life of a customer
 A firm with a 70% customer retention rate has to replace about half of its
customers about every three years to maintain its total number of customers.

Some Additional Thoughts
 When estimating CLV, managers are typically confronted with several

questions:
 CLV models typically assume an infinite time horizon. However, in some cases, it
may be more appropriate to calculate CLV with finite periods.
 The CLV models presented in this chapter assume that firms make, on average,
the same gross margin with a customer each time period once they’re acquired.
Hence, gross margin may need to be adjusted for each time period based on the
expected change in gross margin over time.
 Determining an appropriate retention rate is non-trivial. Similar to the previous
point on using an average gross margin, we need to ask the question of whether
the retention rate for each customer changes over time?
 Managers can calculate CLV using whichever time period is most appropriate.
So, if on average customers purchase once each month, it is worthwhile to
consider calculating CLV at a monthly or at most quarterly level.
 Different CLV formulas and symbols are used in various CLV textbooks. However,
the underlying models are largely the same.

Agenda
 Introduction
 CLV in Action

 An Example
 Summary
 Takeaways

Summary
 Whereas traditional customer measures are often backward looking (i.e.,

they measure the past profitability of a customer relationship), Customer
Lifetime Value (CLV) is a forward-looking approach that enables
researchers to estimate the predicted future value and profitability of a
customer.
 CLV can help improve many marketing decisions, such as which

customers to target, how much to spend to acquire and retain new
customers, and when to stop serving (i.e., “firing”) a customer.

Summary
 CLV also allows firms to compare the costs to acquire a customer with the
future profits expected from that customer. Importantly, CLV can be easily
expanded from an individual customer to a segment of customers, thus
enabling firms to make better targeting decisions (see Chapters 3 & 4
where we discuss segmentation and targeting).
 Because CLV is forward-looking, it requires researchers to make certain

assumptions (e.g., how long will a customer be retained) and hence can be
inaccurate.
 Nevertheless, many firms today recognize the importance and value of

estimating CLV, partially because it has been shown to predict overall firm
value.

Agenda
 Introduction
 CLV in Action

 An Example
 Summary
 Takeaways

Takeaways
 Some customers are more valuable (i.e., profitable) than others.
 Customer Lifetime Value (CLV) is a forward-looking approach that allows

researchers to estimate the value and profitability of customers.
 CLV is the present value of the future cash flows that can be attributed to
a particular customer.
 CLV focuses on customer profitability over the long-term and helps firms
distinguish between profitable and unprofitable customers.

Takeaways
 CLV helps determine the maximum amount firms should spend to

acquire a new customer and/or retain and serve existing customers.
 CLV calculations are forward-looking and rely on assumptions (e.g., how

long will a customer be retained). Thus, CLV can be inaccurate.
Nevertheless, many firms recognize the value of CLV.
 Research has shown that a firm’s CLV-based customer valuation is highly

related with that firm’s market value, in support of the managerial
importance of CLV.
 Although the formulas presented and used in this chapter are the most
commonly used ones, different researchers use different CLV formulas
and/or symbols. Yet, the various formulas are all related and serve the
same purpose.

Chapter 9

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter 9

Uploaded by

Copyright:

Available Formats

Marketing Analytics

Based on First Principles :

Using Customer Lifetime Value

© Palmatier, Petersen, and Germann 1

 Customer Lifetime Value (CLV)

© Palmatier, Petersen, and Germann 2

 Know the idea behind Customer Lifetime Value (CLV).

 Understand the “standard” CLV formula as well as the “alternative” CLV

 Describe how to use CLV for prospecting decisions.

 Know how to use CLV to determine maximum acquisition costs.

 Explain how to do a return on marketing investment (ROMI) calculation

 Understand the different approaches researchers take to estimate the

 Know how CLV is linked to firm value.

© Palmatier, Petersen, and Germann 3

 Customer Lifetime Value (CLV)

© Palmatier, Petersen, and Germann 4

Not all customers are equally profitable.

 Usually about 10-20% of your customers help to generate about 80-90% of

 Further, the next 40-60% of your customers tend to be a little above

© Palmatier, Petersen, and Germann 5

 If a company cannot turn unprofitable customers into profitable ones, the

 Some suggest companies should fire unprofitable customers. Doing so can

© Palmatier, Petersen, and Germann 6

 Consider the following example:

 However, it can be possible to end relationships with customers and come

© Palmatier, Petersen, and Germann 7

 The idea behind CLV is fairly intuitive.

© Palmatier, Petersen, and Germann 8

 As customers, most of us have individual CLV scores with at least some

© Palmatier, Petersen, and Germann 9

 Customer Lifetime Value (CLV)

© Palmatier, Petersen, and Germann 10

Gross Margin ($M) = Price – Unit Variable Cost

Asher $M = $500 – $150 = $350

© Palmatier, Petersen, and Germann 11

+ = $350 + $350 + $350 .....

Year 1 Year 2 Year 3

Year 1 Year 2 Year 3

© Palmatier, Petersen, and Germann 12

Year 1 Year 2 Year 3

© Palmatier, Petersen, and Germann 13

Year 1 Year 2 Year 3

© Palmatier, Petersen, and Germann 14

© Palmatier, Petersen, and Germann 15

Year 1 Year 2 Year 3

 Using this alternative CLV formula, Nick’s CLV is:

© Palmatier, Petersen, and Germann 16

© Palmatier, Petersen, and Germann 17

© Palmatier, Petersen, and Germann 18

 Hence, a prospect’s lifetime value (PLV) can be calculated as follows:

 Let us assume Asher’s Marketers are considering spending $500,000 on a new

 Thus, the PLV of each of the 1,000,000 viewers is:

PLV = 0.001 x $1,925 - $0.5 = $1.925 - $0.5 = $1.425 ≈ $1.43

© Palmatier, Petersen, and Germann 20

 Thus, as long as Asher’s Marketers believe more than 0.026% of the

 That said, measuring the incremental value provided by a campaign can be

 Calculating CLV requires an estimate of each customer’s retention rate. So

 In short, firms use many different approaches to estimate retention rate.

 As another example, some managers may calculate the probability of a

© Palmatier, Petersen, and Germann 22

where P(alive) is the probability of the customer making a purchase from

t is calculated as 10/12 = 0.833

© Palmatier, Petersen, and Germann 24

© Palmatier, Petersen, and Germann 25